Deadtime problem and a temporary solution

NEXT
Make New Entry, Make Followup Entry

User name Kalyan

Log entry time 01:41:25 on December 05, 2008

Entry number 252093

This entry is a followup to: 252087

Followups:

keyword=Deadtime problem and a temporary solution

I was paged to look into the huge negative deadtime problem that shift
crew noticed during run 4186. After talking to Bob and doing some simple
tests we found that it was due to T3 prescale factor not working right.
We called for controlled access and tried to reproduce the problem again
with a pulser plugged into T3.

Here is the symptom:

1. If I plug 1kHz pulser in T8 and set prescale ps8=5 and everything else
infinity, I see ~200Hz(expected) of CODA rate and DT from datamon 1%.

2. For 1.5KHz pulser in T3 with ps3=5 and everything else infinity, I see
~650Hz CODA(not expected) rate and DT from datamon is larger number -ve
no.( like -170, I think).

3. If I do same thing as step 2 for T1, I see ~300Hz CODA rate and DT = 1%

This tells us that the prescale bit for T3 could be problematic and it
could be a part of much bigger problem, as Bob mentioned earlier(?).

To fix this temporarily we swapped T2 with T3 at the trigger supervisor
and for the scaler(both LHRS and BB). As a consequence all T1 and T3
signals will be swapped at the TDC too. This will affect production runs
starting from 4205.

If you see bunch of TDC spectra missing in T3 then don't panic ..look in T2!